Title :
High-throughput screening of DeNOx catalyst using artificial neural networks
Author :
Song Hwa Chae ; Sang Hun Kim ; Park, Sunwon
Author_Institution :
Dept. of Chem. & Biomolecular Eng., Korea Adv. Inst. of Sci. & Technol., Seoul
Abstract :
The support vector regression is used to model the relationship between the inputs (material composition and reaction temperature) and the output (NO conversion). Machine learning algorithms discover the relationships between the variables of a system (input, output and hidden) from direct samples of the system. Usually there is relatively small number of samples compared with the number of input features. Relatively small number of samples and large number of features would cause overfitting. The support vector machine (SVM) avoids overfitting by choosing a specific hyperplane among the many that can separate the data in the feature space. SVM realizes the structural risk minimization. In this study, the support vector machine is applied to predict catalytic activity of various libraries in a quaternary system of Pt, Cu, Fe, and Co supported on aluminium-containing SBA-15 using a self made 64-channel micro reactor. This method would help to discover the optimum composition of DeNOx catalysts
Keywords :
catalysts; chemistry computing; learning (artificial intelligence); neural nets; regression analysis; support vector machines; DeNOx catalyst; SVM; artificial neural network; high-throughput screening; machine learning algorithm; structural risk minimization; support vector machine; support vector regression; Artificial neural networks; Chemical engineering; Chemical technology; Composite materials; Data mining; Electronic mail; Materials science and technology; Risk management; Support vector machines; Throughput; DeNOx catalyst; Support vector machine; high-throughput screening;
Conference_Titel :
SICE-ICASE, 2006. International Joint Conference
Conference_Location :
Busan
Print_ISBN :
89-950038-4-7
Electronic_ISBN :
89-950038-5-5
DOI :
10.1109/SICE.2006.314627
